sign in sign up
<<
Home , Anamalai Tiger Reserve, Western Ghats

bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1 Land Use Change Increases Wildlife Parasite Diversity in Anamalai Hills, Western , 2

3 Debapriyo Chakraborty1,2, D. Mahender Reddy1, Sunil Tiwari1, Govindhaswamy 4 Umapathy1*

5 1 CSIR-Laboratory for the Conservation of Endangered , Centre for Cellular and 6 Molecular Biology, 500048, India 7 2 Present address: EP57 P C Ghosh Road, Kolkata 700048, India 8 *[email protected]

9

10 ABSTRACT

11 Anthropogenic landscape change such as land use change and habitat fragmentation are 12 known to alter wildlife diversity. Since host and parasite diversities are strongly connected, 13 landscape changes are also likely to change wildlife parasite diversity with implication for 14 wildlife health. However, research linking anthropogenic landscape change and wildlife 15 parasite diversity is limited, especially comparing effects of land use change and habitat 16 fragmentation, which often cooccur but may affect parasite diversity substantially 17 differently. Here, we assessed how anthropogenic land use change (presence of , 18 livestock foraging and human settlement) and habitat fragmentation may change the 19 gastrointestinal parasite diversity of wild mammalian host species (n=23) in Anamalai hills, 20 India. We found that presence of , and potentially livestock, significantly 21 increased parasite diversity due possibly to spillover of parasites from livestock to wildlife. 22 However, effect of habitat fragmentation on parasite diversity was not significant. Together, 23 our results showed how human activities may increase wildlife parasite diversity within 24 human-dominated landscape and highlighted the complex pattern of parasite diversity 25 distribution as a result of cooccurrence of multiple anthropogenic landscape changes.

26

27 Keywords: Parasite diversity, Land use change, Plantation, Livestock, Human settlement, 28 Anamalai hills, Habitat fragmentation, Rainforest

29

1 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

30 INTRODUCTION:

31 Land use change and habitat fragmentation are two major landscape-level outcomes of 32 human activities that significantly impact 1–3. Consequently, considerable 33 research on biodiversity change in human-dominated landscape have been conducted, 34 which has resulted in improved understanding of how these two human impacts on 35 landscape can impact biodiversity 1,4,5. These anthropogenic factors can also modify host– 36 parasite interactions, which, in turn, can lead to either increase or decrease in parasite 37 diversity 6–8,8. Understanding how these factors may influence parasite diversity is 38 ecologically important for multiple reasons. For instance, parasites regulate host population 39 dynamics 9, alter species communities 10 and constitute a significant proportion of total 40 biomass of any ecosystem 11, which is not surprising considering parasites comprise at least 41 40% of all animal species on earth 12. Despite their ecological importance, our knowledge on 42 parasite diversity is limited13,14, particularly in the context of increasing human impact on 43 environment, underlining a significant research gap15,16. The gap is specifically wide for 44 wildlife hosts and urgent research is required in the face of recent increased emergence of 45 novel pathogens of wildlife origin 7,17,18. It is, thus, crucial to answer how anthropogenic land 46 use change and habitat fragmentation may impact parasite diversity in the wild.

47 Land use change can affect parasites both directly and indirectly. By altering 48 environment (for example, through pollution), land use change may render transmission of 49 environmentally-transmitted parasites difficult. This is particularly true for parasites that has 50 life stages outside host body. However, land use change can indirectly impact parasite 51 diversity by altering host diversity as it is one of the strongest predictors of parasite diversity 52 19–22. By decreasing host diversity and abundance, land use change can deplete richness of 53 parasites particularly those that require multiple obligatory hosts 23. This is evident when 54 many host species that are threatened in their natural habitat appear to harbour fewer 55 parasites 24. On the other hand, land use change can also increase parasite diversity in 56 multiple ways. Land use change can increase parasite diversity by increasing host diversity. 57 For instance, land use change such as agricultural field or land-fill can act as resource traps 58 and amplify host diversity artificially 25. Land use change can also increase parasite diversity 59 by introducing non-native parasites such as parasites of domestic and feral animals and 60 even from humans 26.

2 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

61 It is also important to distinguish between different types of land use change and 62 their effects on parasites 27. One type of land use change that has not been studied well is 63 the effect of plantation on wildlife parasites 27. Plantations are usually monocultures of 64 exotic or native plant species grown as timber or fuel wood or as cash crops and have a 65 large and increasing footprint in wildlife habitats worldwide 28. They can sometime act as 66 refuge to wildlife but usually with a biotic homogenising effect29,30. Consequently, plantation 67 may also increase but homogenise parasite community. Plantations are often accompanied 68 by settlement of labourers and livestock foraging 31,32. These changes within a wildlife 69 habitat can both increase or decrease parasite diversity. Parasite diversity may decrease if 70 wildlife hosts avoid human areas to lessen confrontation with humans and resource 71 competition with livestock. On the other hand, generalist species may actually thrive in 72 human settlements by utilizing novel resources 33,34. Herbivore species may also prefer to 73 stay closer to human settlements and livestock (“spatial refugia”) that may displace 74 predators 35–38. Moreover, many wildlife, over time, may actually get habituated to humans 75 and livestock and aggregate near human-dominated landscape 39,40. These aggregations may 76 eventually increase parasite diversity by increasing contact between native host species. 77 Such situations may also increasingly expose wildlife to humans and human-associated 78 animals such as livestock and commensals, increasing chance of spillover of non-native 79 parasites to wildlife.

80 Habitat fragmentation may lead to higher parasite diversity because heavily 81 fragmented habitats may disrupt wildlife dispersal and increase host diversity in smaller 82 fragments. Such increase in host diversity in a smaller patch may alter host characteristics 83 such as home range, abundance and intra and interspecific contacts thus increasing overlap 84 among host species making host individuals exposed to higher parasite infections 41,42. 85 These effects are likely to be greatest in the smallest and most isolated of the fragments 3,43. 86 By disrupting host dispersal, fragmentations can also adversely affect parasite diversity. This 87 could be especially true for parasites who require multiple host species to complete its life 88 cycle, such as those that are transmitted trophically 44. So far, many studies looked into this 89 effect but the results have been mixed 6,41,45–47.

90 The Anamalai ( hills in Tamil) hills of southern India is a highly biodiverse 91 rainforest habitat of , which holds about 30% of India's plant and vertebrate

3 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

92 species diversity in less than 6% of the country’s area 48. It is also one of the most altered 93 natural habitats in India and typifies different levels of land use change and habitat 94 fragmentation rampant in Indian wildlife habitats. Large section of the habitat is highly 95 modified due to land use change, bordered by large, relatively undisturbed tropical 96 rainforests. The landscape is a matrix of over 40 rainforest fragments (1-2,500 ha in 97 size),often surrounded by plantations (coffee, tea and cardamom), roads, hydroelectric 98 dams and settlements 49. Highly-modified fragments contain within them human 99 settlements and have higher livestock pressures than other remote, less disturbed 100 fragments. In spite of such high levels of land use change and habitat fragmentation, the 101 Anamalai hills still harbour a large number of wildlife whose ranges often unavoidably 102 overlap with humans and livestock 50–53. In fact, large number of wildlife species are 103 regularly observed within human-dominated habitats and this concurrence with humans 104 often precipitates into wildlife-human conflicts 49,50,54–56. It is possible that many of the 105 wildlife are important reservoirs of multiple environmentally-transmitted parasites. In fact, 106 recent studies have recorded important parasite groups within certain host species 107 populations that may cause Ascariasis, Trichuriasis and Strongylodiasis in humans 45,46,57,58.

108 To assess the effect of land use change (plantation, livestock foraging and human 109 settlements) and habitat fragmentation on parasite diversity, we studied gastrointestinal 110 parasites of wild mammalian hosts across rainforest fragments in Anamalai hills. Using 111 statistical models, we tested effects of land use change and habitat fragmentation on 112 parasite diversity. We predicted a positive impact of land use change on parasite diversity 113 due to increased host diversity and an increased exposure of wildlife to humans and 114 livestock. For habitat fragmentation too, we predicted an increase in parasite diversity with 115 decrease in habitat size and increase in habitat isolation. Our alternative predictions were 116 that land use change and habitat fragmentation could actually deplete parasite diversity by 117 decreasing host diversity in disturbed fragments. Finally, land use change and habitat 118 fragmentation may not significantly impact parasite diversity either by not impacting host 119 community or by not spillover from non-native hosts such as livestock and humans.

4 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

120 MATERIALS AND METHODS:

121 Ethical statement: For this study, faecal samples were collected only noninvasively. As a 122 result, no animal was sacrificed or harmed during sampling. Part of the sampling was done 123 within Anamalai Reserve, which is a protected area. Hence, appropriate written 124 permission was taken from the Forest Department (Letter Ref. No. WL 125 5/58890/2008, dated 2nd September 2009).

126 Study site: Located south of the Palghat gap (11° N) of the Western Ghats, Anamalai hillss 127 once had large tracts of dotted with few tribal settlements. Between 128 1860 and 1930, British colonisers started clearing the rainforests extensively for cultivation 129 of tea and coffee and developing and Eucalyptus plantations, particularly in the 130 Plateau 59. As a result, the Anamalais today consists of both a relatively 131 undisturbed, large (958.59 km2) tropical rainforest within the protected Anamalai Tiger 132 Reserve (ATR; 10°12′–10°35′N and 76°49′–77°24′E) and about 1,000 ha highly degraded 133 Valparai Plateau (Figure 1). The plateau consists of many tea estates and other plantations, 134 which are surrounded by four protected areas—ATR in Tamil Nadu state and three others in 135 state. The major vegetation types include scrub forests in the rain-shadow areas in 136 the eastern foothills, dry and moist deciduous forests (<800 m), mid-elevation tropical wet 137 evergreen forest (600-1,500 m) and high-altitude -grassland ecosystems (>1,500 m) 60. 138 Although a large part of the tropical wet evergreen forests occurs within ATR, many of the 139 smaller (< 200 ha) fragments are found in private estates in the Valparai plateau. These 140 small fragments are highly degraded and disturbed due to fuel-wood collection and 141 livestock grazing. Valparai town also is a part of the plateau and around 200,000 people live 142 across the town and plantations 60. Due to the ongoing habitat fragmentation, the whole 143 landscape is a matrix of over 40 rainforest fragments, ranging 1 ha-2,500 ha in size and 144 often surrounded by plantations (coffee, tea and cardamom), roads, hydroelectric dams and 145 settlements 49. Based on size range (2-2,500 ha), level of perceived human disturbance and 146 access, we selected 19 mid-elevation tropical rainforest and three low-elevation dry and 147 moist deciduous forest fragments for sampling (Figure 1).

148 Host sampling: Between Oct 2013 and Oct 2015, faecal samples were collected from 149 populations of mammalian wildlife. We collected fresh faecal samples, non-invasively during 150 the day, on transects (400 m-3 km in length). For large and medium herbivores and

5 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

151 primates, we followed individuals and collected fresh faeces when animals defaecated. For 152 elusive species such as carnivores, we identified home range based on secondary 153 information and faecal samples were identified based on morphology and also using nearby 154 secondary signs such as pug-marks or hoof-prints. To avoid sampling the same individual 155 repeatedly, only one sample of a host species was collected from each spot and the sample 156 source was either marked or removed. To avoid contamination from soil, samples were 157 collected from the inside of the bolus or only top pellet was collected from a heap. We 158 immediately fixed each sample in 10% formaldehyde solution (50 ml), labelled the 159 containers with the information of origin (fragment name, date, Time and host species) and 160 stored them at room temperature until parasitological screening. Differences in sampling 161 effort can confound the comparison of diversity among replicates. We accounted for 162 differences in number of host species encountered by calculating richness estimates with 163 the assumption that each faecal sample represents single individual. We used bootstrap, 164 which is a resampling method for estimating the whole sampling distribution of richness by 165 sampling with replacement from the original sample and can offer greater precision than 166 jackknife estimates, especially when sample sizes are small 61.

167 Parasite sampling: Employing both the flotation and sedimentation techniques (NaNO3 168 solution), we screened the faecal samples for the presence of helminth eggs, larvae and 169 protozoan cysts 62. For each parasite concentration technique, we examined two slides per 170 sample. Slides were examined under a light microscope (400X). Eggs and cysts were first 171 examined at 10× magnification and then their size was measured with a micrometre 172 eyepiece (0.1 μm) at 40× magnification. To facilitate identification of parasite eggs, we often 173 added a drop of Lugol’s iodine solution to the slides, which highlighted detailed structures. 174 In addition, photographs of each parasite species have been archived and are available for 175 examination by request to the corresponding author. We identified parasites to the lowest 176 possible taxonomic level using published keys 63,64. Differences in sampling effort can 177 confound the comparison of diversity among replicates. We accounted for differences in 178 number of parasite taxon encountered by calculating richness estimates with the 179 assumption that each faecal sample represents single host individual. We used bootstrap, 180 which is a resampling method for estimating the whole sampling distribution of richness by 181 sampling with replacement from the original sample. Bootstrap can offer greater precision

6 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

182 over jackknife estimator, especially when sample sizes are small 61. This method is 183 particularly recommended for parasite richness estimation 65.

184 Land use data: In Anamalai hills, land use change manifests in largely three forms— 185 presence of human settlements, plantations and livestock foraging. There are only few large 186 (>1000 ha) fragments that are legally protected and thus undisturbed. Many of the studied 187 fragments share more than one type of land use change. For instance, some fragments with 188 human settlements may also have livestock present. For the current study, we identified 18 189 fragments with land use change, out of which 15 (83.3%) had plantation, in contrast to three 190 (16.7%). Eleven (61.1%) of the fragments have significant livestock foraging pressure, in 191 contrast to seven (38.9%) fragments without livestock. Finally, ten (55.6%) of the fragments 192 had human settlements within them, in contrast to eight (44.4%) without settlements.

193 Habitat fragmentation data: To measure effect of habitat fragmentation, we used fragment 194 size and isolation distance between fragments. According to the equilibrium theory of island 195 biogeography, organism dispersal probability declines as distance between islands 196 increases, reducing rates of immigration and, in turn, reducing diversity (MacArthur & 197 Wilson, 1963, 1967; Whittaker & Fernandez-Palacios, 2007). Assuming each forest fragment 198 as an island, their isolation was summarized with an isolation index which was calculated as 199 the sum of the square root of the distances to the nearest equivalent (no smaller than 80% 200 of size) or larger fragment (Dahl, 2004). Data on fragment size, distance between fragments 201 and presence of human settlements, plantations and livestock were collected from earlier 202 studies from Anamalai hills 45,60.

203 Data analyses: To assess the effects of land use change and habitat fragmentations on 204 bootstrap estimate of parasite taxon richness, we created two different linear mixed effects 205 models 66. Each model included random effects of host species and fragments to account for 206 multiple observations within each fragment (across host species) and across fragments. In 207 the land use model, the predictor variables (fixed effects) were presence of plantation, 208 human settlement and livestock. The predictor variables for the habitat fragmentation 209 model were fragment size and fragment isolation index. In both the models, we 210 incorporated both bootstrap estimates of host and host body mass as co- 211 predictors as these were known to effect parasite richness. We retrieved host body mass 212 data from online ecological database 67. After fitting these model to the data, we also

7 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

213 compared and selected the best fit model using lowest AIC value 68. At the end, diagnostics 214 were run to check distribution of the residuals for each model. This analysis was conducted 215 in the lme4 package 69. We also assessed the effects of land use change and habitat 216 fragmentation on bootstrap estimates of host species richness using two linear models. In 217 the land use model, the predictor variables were presence of plantation, human settlement 218 and livestock. The predictor variables for the habitat fragmentation model were fragment 219 size and fragment isolation index. We followed the same strategy as described above for 220 model fitting, fitting diagnostics and model selection. Finally, we tested whether land use 221 change homogenized the composition of the parasite community. We used a multivariate 222 nonparametric Analysis of Variance (permAnoVa; 1,000 permutations) based on the Jaccard 223 dissimilarity index for a matrix of parasite presence/absence. We calculated the variance of 224 homogeneity of parasite communities within each fragment based on disturbed vs. 225 undisturbed divisions using the betadisper function of the vegan package in R 70.

226

227 RESULTS:

228 Sample diversity: From 19 forest fragments, we collected 4,056 mammalian faecal samples 229 belonging to 23 mammalian wildlife species and two livestock species—domestic goats 230 (Capra aegagrus) and cattle (Bos taurus). Analyses were done only on wildlife samples. 231 Number of samples varied from 41 in Uralikal to 495 in Puthuthottam (Table 1). Number of 232 samples for each host species varied between six in Otter (Lutra lutra) and 623 in (B. 233 gaurus). In total, seven protozoa (18.42%) and 32 helminth (81.58%) species were recorded, 234 including five trematodes, five cestodes and 20 nematodes. At least seven different 235 parasites, belonging to different parasite groups, were recorded in ≥20 different host 236 species—protozoa Coccidia sp. (23 hosts); cestodes Hymenolepis nana (20 hosts) and 237 Moniezia sp. (22 hosts); and nematodes Gongylonema sp. (20 hosts), Strongyloides sp. (23 238 hosts), Trichuris sp. (24 hosts) and Ascaris sp. (26 hosts). On the other hand, cestode 239 Dipylidium sp. and nematode Parascaris sp. were found only in civet and Indian porcupine 240 (Hystrix indica) samples, respectively.

241 Host and parasite diversity and disturbance: For parasite diversity analysis the human 242 disturbance model was the best fit (Table 2). Parasite diversity was significantly driven by

8 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

243 presence of plantation (estimate = 4.779, CIProfile = 0.326 – 9.232, t = 2.103, p<0.05). 244 Presence of livestock had a substantial but not significant positive effect (estimate = 3.209,

245 CIProfile = -0.052 - 6.366, t = 1.992, p>0.05). Effects of settlement, host richness and host body 246 mass on parasite richness were not significant (Figure 2). For host diversity analysis the 247 human disturbance model was again the best fit (Table 3). Presence of plantation was the 248 only predictor that had a significant positive effect on host diversity (estimate = 10.798,

249 CIProfile = 2.302 - 19.294, t = 2.726, p<0.05)—almost half of all host species occur in 250 plantations. Although presence of livestock did not have a significant effect, its wide 251 confidence interval was mostly on the positive side suggesting potential positive impact—

252 limited by sample size—on host richness (estimate = 5.602, CIProfile = -0.639- 11.843, t = 253 1.925, p>0.05). Similarly, presence of human settlement did not significantly affect host 254 richness, however, the substantial effect was mostly on the negative side, suggesting

255 potential negative effect on host diversity (estimate = -4.112, CIProfile = -10.717- 2.492, t = - 256 1.335, p>0.05).

257 We recorded 12 parasites (ten helminths and two protozoa) that occurred only in 258 plantations. Six of the ten helminths were nematodes (60%), while rest were trematodes 259 (30%) and one cestode (10%). Fragments without plantations did not harbour any parasite 260 taxon exclusively, which means parasites in those undisturbed fragments also occured in 261 plantations. Fragments with livestock harboured three parasite taxa (two nematodes and 262 one cestode) exclusively relative to their undisturbed counterpart. However, only one 263 parasite taxon (Taenia sp.) exclusively occurred in livestock disturbed fragments, while other 264 two nematodes also occurred in the plantations. Its counterpart undisturbed fragments only 265 harboured one taxon exclusively (Paragonimus sp.), which however also occurred in 266 plantations. Finally, settlements harboured three nematode taxa exclusively in comparison 267 to their undisturbed counterpart. Only one of these taxa (Uncinaria sp.) were exclusive to 268 settlements across all fragments. Undisturbed counterpart of settlements harboured only 269 one parasite taxon (Sarcocystis sp.) exclusively.

270 Parasite and host homogeneity: Parasite communities within disturbed forest fragments 271 were not significantly more homogeneous than the undisturbed ones due to presence of 272 either plantations (F = 2.58, p>0.05), livestock (F = 0.04, p>0.05) or settlements (F = 3.55, 273 p>0.05). Host communities within plantations (TukeyHSD; p<0.05; Figure 4a) and human

9 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

274 settlements (TukeyHSD; p<0.05; Figure 4b) were, however, significantly more homogeneous 275 than undisturbed fragments. Finally, we did not find any of the disturbance variables to 276 significantly alter the parasite community composition between undisturbed and disturbed 277 fragments.

278

279 DISCUSSION:

280 Our results reveal that rainforest fragments with plantations (and potentially with livestock) 281 in Anamalai hills harbour significantly higher parasite diversity than undisturbed fragments. 282 Interestingly, some of the modified fragments (at least, fragments with plantations) also has 283 significantly more host diversity than the undisturbed fragments, however host diversity 284 was not found to significantly affect parasite diversity.

285 In Anamalai hills, plantations (coffee, tea and cardamom) had more mammalian 286 wildlife species richness than the undisturbed fragments. This was not particularly a 287 surprising result because studies have reported similar high richness in vertebrate species 288 from plantation within wildlife habitats 71,72. In fact, earlier studies from Western Ghats also 289 found high vertebrate richness within or around plantations with large variations depending 290 on plantation types, from open tea to more shaded coffee and cardamom plantations 291 30,73,74. The reason for such increased host diversity is thought to be an increase in habitat 292 heterogeneity within plantations. Increased habitat heterogeneity is thought to generate 293 greater diversity of niches consequently facilitating cooccurrence of many species 75,76. 294 However, such increase in species richness is often accompanied by more generalist and 295 wide-ranging species being more abundant within the plantations and a loss of community 296 heterogeneity relative to undisturbed habitats 77–79. We found similar loss of heterogeneity 297 for host species in disturbed habitats with plantations and settlements (Figure 4).

298 Effect of livestock presence on host species richness was positive but not statistically 299 significant at α = 0.05. The effect, however, was significant at α = 0.10, which suggested 300 potential, but weak effect, which was reflected by the almost equal number of wildlife

301 species recorded from these two groups of fragments (nLivestock = 20 and nUndisturbed = 22). 302 Interaction between livestock and wildlife is complicated. For instance, while a number of 303 studies found evidence of competitive exclusion between livestock and large herbivore 80,

10 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

304 may other recorded resource sharing between these two groups 81,82. Yet still, may other 305 studies did not find any relationship between the two 83. The outcome of the interaction 306 may depend on the ecological similarity between the two groups (Niche overlap), 307 availability of natural resources that may vary between habitats (between low to high 308 productivity) and also degree of behavioural habituation by the wildlife. The wildlife 309 community that we studied was an ecologically broad one consisting of wildlife with very 310 different ecology. Therefore, while some of the species—such as spotted deer and sambar 311 deer, who were found only in the undisturbed fragments—may face resource competition 312 from livestock grazing, others (for example, small carnivores and primates) may not face any 313 competition. In addition, many large herbivores, such as gaur and , who despite 314 resource competition may still use the disturbed fragments as corridors contributing to host 315 richness. These processes together may explain almost similar host species richness 316 between fragments with and without livestock grazing.

317 We did not find any significant effect of human settlement on host diversity but the 318 trend is negative (Figure 3). While human settlement may attract and facilitate generalist 319 and weedy species with high tolerance for disturbance (for example, rodents, which were 320 not sampled in the present study), many elusive species such as carnivores may be 321 adversely affected and may prefer to avoid fragments with settlements 84. Still, we recorded

322 overall a large host species richness (host richnessSettlement = 19, host richnessUndisturbed = 23) 323 from around the settlement in Anamalai hills. This could be explained by the facts that many 324 of these settlements may attract wildlife with unintentionally supplemented resources such 325 as planted fruit trees 60. Additionally, the high level of fragmentation of the landscape 326 meant large herbivores and carnivores may not have much choice but to disperse through 327 human settlements 54,56. We did not find evidence of habitat fragmentation (fragment size 328 and isolation) influencing host species richness in Anamalai hills (Figure 3). This is in line with 329 findings from across studies that effects of fragmentation on species communities are often 330 weak 85. Effects of habitat fragmentation on species diversity is highly context-specific and 331 varies considerably between animal groups, ecosystems and kinds of human activities 332 prevalent in the landscape 85–89. In Anamalai hills, habitat fragmentation is widespread, 333 which likely disrupt animal movement to some extent but, in the absence of hunting, 334 perhaps not substantially. For instance, studies recorded use of certain plantation as

11 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

335 corridors to connect with isolated undisturbed habits 51,52,74. However, the adverse outcome 336 of these movement through human-dominated habitats is the increase in wildlife-human 337 conflict 54,56.

338 Among the different types of land use change in Anamalai hills, plantations had the 339 strongest positive effect on parasite diversity (Figure. 2). Increase in number of parasite taxa 340 in modified fragments ranged between one to ten, with eight parasite taxa that were 341 recorded exclusively in these fragments (Table 4). However, this increased richness in 342 disturbed fragments were likely not driven by host richness as host richness had a small and 343 statistically not significant effect on parasite richness in the disturbance model (Figure 2). 344 This is in contrast to the predominant patterns across most studies on parasite diversity that 345 found host richness to be the strongest predictor of parasite richness 19–22. However, there 346 could be potential deviations from this rule, particularly due to human impacts 21,90,91. For 347 instance, many human parasites may spillover to wildlife (anthropozoonoses) as humans 348 regularly come in contact with wildlife 92–97. Humans may also introduce many non-wildlife 349 species such as feral dogs, cats in addition to livestock into wildlife habitats and these 350 species may share parasites with wildlife 26. In such cases, parasite richness in wildlife would 351 be more than in the undisturbed fragments. Indeed, all but one (Schistosoma sp.) of the 352 parasites that we found exclusively in plantations also occurred in cattle (Table 4). 353 Surprisingly, wildlife parasite taxa that were present in the livestock foraging fragments did 354 not occur in cattle samples from the same fragments. This was also the case for the wildlife 355 parasites that only occurred in settlement but not in undisturbed fragments. We did not find 356 any significant effect of host body mass on parasite diversity (Figure 2). This is in contrast 357 many studies that found a significant relationship between these two variables 98,99. On the 358 other hand, many other empirical studies that did not find any relationship between body 359 mass and parasite richness when accounting for host phylogenetic relationships 100,101. Such 360 contradictory results may suggest that relationship between host body mass and parasite 361 diversity is a factor of body mass and life history traits, which vary between ecologically 362 different groups of hosts 22. Thus, the broad ecological diversity among host species in the 363 present study might have confounded this relationship.

364 Our results did not find any significant effect of habitat fragmentation on parasite 365 richness (Figure 2). This was expected as we did not find any effect of fragmentation on host

12 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

366 richness either. This lack of relationship between fragmentation and parasite diversity could 367 also be an outcome of large home ranges and low habitat specialisation of most of the host 368 species in our study. Many of the species that we sampled were large herbivores or 369 carnivores (e.g., Elephas maximus, Bos gaurus, Panthera tigris, Panthera pardus) with lang 370 home range and they disperse across fragments. The level of fragment isolation (Median 371 distance = 30.2 km) may not be a deterrent to their dispersal. Similarly, many host species in 372 the study community such as Macaca radiata, Sus scrofa, Viverricula indica, are habitat 373 generalists. According to the distribution-abundance relationship hypothesis 6, smaller, 374 fragmented habitats may be conducive for these generalist wildlife with high reproductive 375 rates. These hosts may then spread parasites across habitats, independent of the level of 376 habitat fragmentation.

377

378 CONCLUSION:

379 With data on 40 parasites from 23 host species collected from 19 rainforest fragments with 380 different types of land use change, we demonstrated that land use changes increased 381 parasite diversity and presence of potential spillover of parasites from livestock to wildlife. 382 We also showed that the observed pattern of parasite diversity was not driven by habitat 383 fragmentation.

384 One of the limitations of this study was that it could not test the effect of land use change 385 and habitat fragmentation on the relationship between host density and parasite diversity. 386 Host density is an important predictor of parasite diversity and in nature, host density is 387 linked to host ecology (e.g., home range). However, land use change can unpredictably 388 change host density, which may have a complex outcome for parasite diversity. It will thus 389 be worthwhile in future to explore this question in the present system. Additionally, with 390 the present evidence of potential anthropozonosis, it will be important in future to compare 391 parasite from the present study to samples from humans, livestock and commensal animals 392 in the fragments. Finally, as far as land use change and habitat fragmentation of wildlife 393 habitats in India are concerned, the present study represents a case study with particular 394 relevance for tropical rainforest habitats. However, there exists a large diversity in habitats 395 and levels of disturbance in India. Given the increased threat to wildlife health from

13 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

396 anthropogenic environmental change, it will thus be crucial for wildlife conservation to 397 study the patterns of parasite diversity in other types of habitats, especially those with 398 already threatened wildlife.

399 Data availability: The datasets generated and analysed during the current study are 400 available from the corresponding author on request.

401 Acknowledgements: DC gratefully acknowledges Steffen Foerster for initial discussions on 402 model selection. DC also used part of Fulbright fellowship to work on the current paper, for 403 which he gratefully acknowledges support of US-India Education Foundation (US-IEF), New 404 Delhi. GU gratefully acknowledges funding awarded to him by Department of 405 Biotechnology, (Ref. No. BTPR4503/BCE/8/899/2012 dated 09-11- 406 2012).

407 Author Contributions: Sampling and data collection were done by DMR, ST and DC. DC 408 conducted data analysis and wrote manuscript. GU conceived, designed and lead the study, 409 organised funding and reviewed the manuscript.

410 Competing interests: The authors declare no competing interests.

411 References:

412 1. Foley, J. A. et al. Global consequences of land use. science 309, 570–574 (2005). 413 2. Goudie, A. S. Human impact on the natural environment. (John Wiley & Sons, 2018). 414 3. Haddad, N. M. et al. Habitat fragmentation and its lasting impact on Earth’s ecosystems. 415 Sci. Adv. 1, e1500052 (2015). 416 4. Dornelas, M. et al. Assemblage time series reveal biodiversity change but not systematic 417 loss. Science 344, 296–299 (2014). 418 5. McGill, B. J., Dornelas, M., Gotelli, N. J. & Magurran, A. E. Fifteen forms of 419 biodiversity trend in the Anthropocene. Trends Ecol. Evol. 30, 104–113 (2015). 420 6. Bordes, F. et al. Habitat fragmentation alters the properties of a host–parasite network: 421 rodents and their helminths in South‐East . J. Anim. Ecol. 84, 1253–1263 (2015). 422 7. Keesing, F. et al. Impacts of biodiversity on the emergence and transmission of infectious 423 diseases. Nature 468, 647 (2010).

14 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

424 8. Wood, C. L., Sandin, S. A., Zgliczynski, B., Guerra, A. S. & Micheli, F. Fishing drives 425 declines in fish parasite diversity and has variable effects on parasite abundance. Ecology 426 95, 1929–1946 (2014). 427 9. Anderson, R. M., Jackson, H. C., May, R. M. & Smith, A. M. Population dynamics of fox 428 rabies in . Nature 289, 765 (1981). 429 10. Wood, C. L. et al. Parasites alter community structure. Proc. Natl. Acad. Sci. 104, 9335– 430 9339 (2007). 431 11. Kuris, A. M. et al. Ecosystem energetic implications of parasite and free-living biomass 432 in three estuaries. Nature 454, 515 (2008). 433 12. Dobson, A., Lafferty, K. D., Kuris, A. M., Hechinger, R. F. & Jetz, W. Homage to 434 Linnaeus: how many parasites? How many hosts? Proc. Natl. Acad. Sci. 105, 11482– 435 11489 (2008). 436 13. Bordes, F. & Morand, S. Parasite diversity: an overlooked metric of parasite pressures? 437 Oikos 118, 801–806 (2009). 438 14. Poulin, R. Parasite biodiversity revisited: frontiers and constraints. Int. J. Parasitol. 44, 439 581–589 (2014). 440 15. Dunn, R. R., Harris, N. C., Colwell, R. K., Koh, L. P. & Sodhi, N. S. The sixth mass 441 coextinction: are most parasites and mutualists? Proc. R. Soc. Lond. 442 B Biol. Sci. 276, 3037–3045 (2009). 443 16. Dunn, R. R., Davies, T. J., Harris, N. C. & Gavin, M. C. Global drivers of human 444 pathogen richness and prevalence. Proc. R. Soc. Lond. B Biol. Sci. 277, 2587–2595 445 (2010). 446 17. Jones, K. E. et al. Global trends in emerging infectious diseases. Nature 451, 990 (2008). 447 18. Smith, K. F. et al. Global rise in human infectious disease outbreaks. J. R. Soc. Interface 448 11, 20140950 (2014). 449 19. Dallas, T. A. et al. Gauging support for macroecological patterns in helminth parasites. 450 Glob. Ecol. Biogeogr. 27, 1437–1447 (2018). 451 20. Kamiya, T., O’dwyer, K., Nakagawa, S. & Poulin, R. Host diversity drives parasite 452 diversity: meta-analytical insights into patterns and causal mechanisms. Ecography 37, 453 689–697 (2014). 454 21. Krasnov, B. R., Shenbrot, G. I., Khokhlova, I. S. & Degen, A. A. Relationship between 455 host diversity and parasite diversity: flea assemblages on small . J. Biogeogr. 456 31, 1857–1866 (2004).

15 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

457 22. Morand, S. (macro-) Evolutionary ecology of parasite diversity: From determinants of 458 parasite species richness to host diversification. Int. J. Parasitol. Parasites Wildl. 4, 80– 459 87 (2015). 460 23. Lafferty, K. D. Environmental parasitology: what can parasites tell us about human 461 impacts on the environment? Parasitol. Today 13, 251–255 (1997). 462 24. Altizer, S., Nunn, C. L. & Lindenfors, P. Do threatened hosts have fewer parasites? A 463 comparative study in primates. J. Anim. Ecol. 76, 304–314 (2007). 464 25. Becker Daniel J., Hall Richard J., Forbes Kristian M., Plowright Raina K. & Altizer 465 Sonia. Anthropogenic resource subsidies and host–parasite dynamics in wildlife. Philos. 466 Trans. R. Soc. B Biol. Sci. 373, 20170086 (2018). 467 26. Weinstein, S. B. & Lafferty, K. D. How do humans affect wildlife nematodes? Trends 468 Parasitol. 31, 222–227 (2015). 469 27. Gottdenker, N. L., Streicker, D. G., Faust, C. L. & Carroll, C. R. Anthropogenic land use 470 change and infectious diseases: a review of the evidence. Ecohealth 11, 619–632 (2014). 471 28. Brockerhoff, E. G., Jactel, H., Parrotta, J. A., Quine, C. P. & Sayer, J. Plantation forests 472 and biodiversity: oxymoron or opportunity? Biodivers. Conserv. 17, 925–951 (2008). 473 29. Solar, R. R. de C. et al. How pervasive is biotic homogenization in human-modified 474 tropical forest landscapes? Ecol. Lett. 18, 1108–1118 (2015). 475 30. Bali, A., Kumar, A. & Krishnaswamy, J. The mammalian communities in coffee 476 plantations around a protected area in the Western Ghats, India. Biol. Conserv. 139, 93– 477 102 (2007). 478 31. Wakker, E., Watch, S. & Rozario, J. de. Greasy palms: the social and ecological impacts 479 of large-scale oil palm plantation development in . Greasy Palms Soc. 480 Ecol. Impacts Large-Scale Oil Palm Plant. Dev. Southeast Asia (2004). 481 32. Richardson, B. Plantation Infrastructure and Labor Mobility in Guiana and Trinidad. 482 Migr. Dev. Implic. For 205–224 (1975). 483 33. McKinney, M. L. & Lockwood, J. L. Biotic homogenization: a few winners replacing 484 many losers in the next mass extinction. Trends Ecol. Evol. 14, 450–453 (1999). 485 34. Sih, A., Ferrari, M. C. & Harris, D. J. Evolution and behavioural responses to human- 486 induced rapid environmental change. Evol. Appl. 4, 367–387 (2011). 487 35. Muhly, T. B., Semeniuk, C., Massolo, A., Hickman, L. & Musiani, M. Human activity 488 helps prey win the predator-prey space race. PLoS One 6, e17050 (2011). 489 36. Berger, J. Fear, human shields and the redistribution of prey and predators in protected 490 areas. Biol. Lett. 3, 620–623 (2007).

16 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

491 37. Hebblewhite, M. et al. Human activity mediates a trophic cascade caused by wolves. 492 Ecology 86, 2135–2144 (2005). 493 38. Ripple, W. J. & Beschta, R. L. Linking a cougar decline, trophic cascade, and 494 catastrophic regime shift in Zion National Park. Biol. Conserv. 133, 397–408 (2006). 495 39. Samia, D. S., Nakagawa, S., Nomura, F., Rangel, T. F. & Blumstein, D. T. Increased 496 tolerance to humans among disturbed wildlife. Nat. Commun. 6, 8877 (2015). 497 40. Blumstein, D. T. Habituation and sensitization: new thoughts about old ideas. Anim. 498 Behav. 120, 255–262 (2016). 499 41. Gillespie, T. R. & Chapman, C. A. Forest fragmentation, the decline of an endangered 500 primate, and changes in host–parasite interactions relative to an unfragmented forest. Am. 501 J. Primatol. Off. J. Am. Soc. Primatol. 70, 222–230 (2008). 502 42. Kowalewski, M. M. et al. Black and gold howler monkeys (Alouatta caraya) as sentinels 503 of ecosystem health: patterns of zoonotic protozoa infection relative to degree of human– 504 primate contact. Am. J. Primatol. 73, 75–83 (2011). 505 43. McCallum, H. & Dobson, A. Disease, habitat fragmentation and conservation. Proc. R. 506 Soc. Lond. B Biol. Sci. 269, 2041–2049 (2002). 507 44. Lafferty, K. D. The evolution of trophic transmission. Parasitol. Today 15, 111–115 508 (1999). 509 45. Chakraborty, D. et al. Mammalian gastrointestinal parasites in rainforest remnants of 510 Anamalai Hills, Western Ghats, India. J. Biosci. 40, 399–406 (2015). 511 46. Hussain, S., Ram, M. S., Kumar, A., , S. & Umapathy, G. Human presence 512 increases parasitic load in endangered lion-tailed (Macaca silenus) in its 513 fragmented rainforest habitats in southern India. PLoS One 8, e63685 (2013). 514 47. Wells, K., Smales, L. R., Kalko, E. K. & Pfeiffer, M. Impact of rain-forest on 515 helminth assemblages in small mammals (Muridae, Tupaiidae) from Borneo. J. Trop. 516 Ecol. 23, 35–43 (2007). 517 48. Mudappa, D. & Raman, T. S. Beyond the borders: Wildlife conservation in landscapes 518 fragmented by plantation crops in India. 21 (Nature Conservation Foundation, 2012). 519 49. Mudappa, D. & Raman, T. S. Rainforest restoration and wildlife conservation on private 520 lands in the Western Ghats. Mak. Conserv. Work 210–240 (2007). 521 50. Singh, M. et al. Distribution, population structure, and conservation of lion-tailed 522 macaques (Macaca silenus) in the , Western Ghats, India. Am. J. 523 Primatol. Off. J. Am. Soc. Primatol. 57, 91–102 (2002).

17 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

524 51. Kumar, M. A., Mudappa, D. & Raman, T. S. Elephas maximus habitat 525 use and ranging in fragmented rainforest and plantations in the Anamalai Hills, India. 526 Trop. Conserv. Sci. 3, 143–158 (2010). 527 52. Navya, R., Athreya, V., Mudappa, D. & Raman, T. S. Assessing occurrence in 528 the plantation landscape of Valparai, Anamalai Hills. Curr. Sci. 107, 1381–1385 (2014). 529 53. Kumaraguru, A., Saravanamuthu, R., Brinda, K. & Asokan, S. Prey preference of large 530 carnivores in , India. Eur. J. Wildl. Res. 57, 627–637 (2011). 531 54. Sidhu, S., Raghunathan, G., Mudappa, D. & Raman, T. R. Conflict to Coexistence: 532 Human-Leopard Interactions in a Plantation Landscape in Anamalai Hills, India. 533 Conserv. Soc. 15, (2017). 534 55. Baskaran, N., Kannan, G., Anbarasan, U., Thapa, A. & Sukumar, R. A landscape-level 535 assessment of Asian elephant habitat, its population and elephant–human conflict in the 536 Anamalai hill ranges of southern Western Ghats, India. Mamm. Biol.-Z. Für Säugetierkd. 537 78, 470–481 (2013). 538 56. Kumar, A. M., Malizia, N. & Koschinsky, J. Spatial Analysis of Human-Elephant 539 Conflicts in a Fragmented Tropical Landscape. (GeoDa Center for Geospatial Analysis 540 and Computation, 2011). 541 57. Chakraborty, D., Tiwari, S., Reddy, D. M. & Umapathy, G. Prevalence of gastrointestinal 542 parasites in civets of fragmented rainforest patches in Anamalai Hills, Western Ghats, 543 India. J. Parasitol. 102, 463–467 (2016). 544 58. Tiwari, S., Reddy, D. M., Pradheeps, M., Sreenivasamurthy, G. S. & Umapathy, G. 545 Prevalence and co-occurrence of gastrointestinal parasites in 546 (Trachypithecus johnii) of fragmented landscape in Anamalai Hills, Western Ghats, 547 India. Curr. Sci. 113, 2194 (2017). 548 59. Congreve, H. The Anamalais. (1942). 549 60. Umapathy, G. & Kumar, A. The occurrence of arboreal mammals in the rain forest 550 fragments in the Anamalai Hills, . Biol. Conserv. 92, 311–319 (2000). 551 61. Hellmann, J. J. & Fowler, G. W. Bias, precision, and accuracy of four measures of 552 species richness. Ecol. Appl. 9, 824–834 (1999). 553 62. Gillespie, T. R. Noninvasive assessment of gastrointestinal parasite infections in free- 554 ranging primates. Int. J. Primatol. 27, 1129 (2006). 555 63. Sloss, M. W., Kemp, R. L. & Zajac, A. M. Veterinary clinical parasitology. 1994. Vet. 556 Clin. Parasitol. Iowa State Univ. Press Google Sch. 557 64. Foreyt, W. J. Veterinary parasitology reference manual. (John Wiley & Sons, 2013).

18 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

558 65. Poulin, R. Comparison of three estimators of species richness in parasite component 559 communities. J. Parasitol. 485–490 (1998). 560 66. Bolker, B. M. et al. Generalized linear mixed models: a practical guide for ecology and 561 evolution. Trends Ecol. Evol. 24, 127–135 (2009). 562 67. Jones, K. E. et al. PanTHERIA: a species-level database of life history, ecology, and 563 geography of extant and recently extinct mammals: Ecological Archives E090-184. 564 Ecology 90, 2648–2648 (2009). 565 68. Aho, K., Derryberry, D. & Peterson, T. Model selection for ecologists: the worldviews of 566 AIC and BIC. Ecology 95, 631–636 (2014). 567 69. Bates, D., Sarkar, D., Bates, M. D. & Matrix, L. The lme4 package. R Package Version 2, 568 74 (2007). 569 70. Oksanen, J. et al. Vegan: Community Ecology Package. R package vegan, vers. 2.2-1; 570 2015. (2018). 571 71. Bhagwat, S. A., Willis, K. J., Birks, H. J. B. & Whittaker, R. J. Agroforestry: a refuge for 572 tropical biodiversity? Trends Ecol. Evol. 23, 261–267 (2008). 573 72. Felton, A., Knight, E., Wood, J., Zammit, C. & Lindenmayer, D. A meta-analysis of 574 and flora species richness and abundance in plantations and pasture lands. Biol. 575 Conserv. 143, 545–554 (2010). 576 73. Ranganathan, J., Daniels, R. R., Chandran, M. S., Ehrlich, P. R. & Daily, G. C. 577 Sustaining biodiversity in ancient tropical countryside. Proc. Natl. Acad. Sci. 105, 578 17852–17854 (2008). 579 74. Sidhu, S., Shankar Raman, T. R. & Goodale, E. Effects of plantations and home-gardens 580 on tropical forest bird communities and mixed-species bird flocks in the southern 581 Western Ghats. J. Bombay Nat. Hist. Soc. 107, 91 (2010). 582 75. MacArthur, R. H. & MacArthur, J. W. On bird species diversity. Ecology 42, 594–598 583 (1961). 584 76. Rosenzweig, M. L. Species diversity in space and time. (Cambridge University Press, 585 1995). 586 77. Harvey, C. A. & Villalobos, J. A. G. Agroforestry systems conserve species-rich but 587 modified assemblages of tropical birds and bats. Biodivers. Conserv. 16, 2257–2292 588 (2007). 589 78. Waltert, M., Bobo, K. S., Sainge, N. M., Fermon, H. & Mühlenberg, M. From forest to 590 farmland: habitat effects on Afrotropical forest bird diversity. Ecol. Appl. 15, 1351–1366 591 (2005).

19 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

592 79. Waltert, M. et al. Assessing conservation values: biodiversity and endemicity in tropical 593 land use systems. PLoS One 6, e16238 (2011). 594 80. Madhusudan, M. D. Recovery of wild large herbivores following livestock decline in a 595 tropical Indian wildlife reserve. J. Appl. Ecol. 41, 858–869 (2004). 596 81. Rannestad, O. T., Danielsen, T., Moe, S. R. & Stokke, S. Adjacent pastoral areas support 597 higher densities of wild ungulates during the wet season than the Lake Mburo National 598 Park in Uganda. J. Trop. Ecol. 22, 675–683 (2006). 599 82. Sitters, J., Heitkönig, I. M., Holmgren, M. & Ojwang, G. S. Herded cattle and wild 600 grazers partition water but share forage resources during dry years in East African 601 savannas. Biol. Conserv. 142, 738–750 (2009). 602 83. Khan, J. A. Conservation and management of Gir lion sanctuary and national park, 603 , India. Biol. Conserv. 73, 183–188 (1995). 604 84. Vanthomme, H., Kolowski, J., Korte, L. & Alonso, A. Distribution of a community of 605 mammals in relation to roads and other human disturbances in Gabon, Central . 606 Conserv. Biol. 27, 281–291 (2013). 607 85. Fahrig, L. Habitat fragmentation: A long and tangled tale. Glob. Ecol. Biogeogr. 28, 33– 608 41 (2019). 609 86. Fahrig, L. et al. Is habitat fragmentation bad for biodiversity? Biol. Conserv. 230, 179– 610 186 (2019). 611 87. Fletcher Jr, R. J. et al. Is habitat fragmentation good for biodiversity? Biol. Conserv. 226, 612 9–15 (2018). 613 88. Simmonds, J. S., van Rensburg, B. J., Tulloch, A. I. & Maron, M. Landscape-specific 614 thresholds in the relationship between species richness and natural land cover. J. Appl. 615 Ecol. (2019). 616 89. Wintle, B. A. et al. Global synthesis of conservation studies reveals the importance of 617 small habitat patches for biodiversity. Proc. Natl. Acad. Sci. 116, 909–914 (2019). 618 90. Wood, C. L. & Johnson, P. T. How does space influence the relationship between host 619 and parasite diversity? J. Parasitol. 102, 485–495 (2016). 620 91. Wood, C. L. et al. Human impacts decouple a fundamental ecological relationship—The 621 positive association between host diversity and parasite diversity. Glob. Change Biol. 24, 622 3666–3679 (2018). 623 92. Eley, R. M., Strum, S. C., Muchemi, G. & Reid, G. D. F. Nutrition, body condition, 624 activity patterns, and parasitism of free-ranging troops of olive baboons (Papio anubis) in 625 Kenya. Am. J. Primatol. 18, 209–219 (1989).

20 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

626 93. Gillespie, T. R. & Chapman, C. A. Prediction of parasite infection dynamics in primate 627 metapopulations based on attributes of forest fragmentation. Conserv. Biol. 20, 441–448 628 (2006). 629 94. Gillespie, T. R. et al. Demographic and ecological effects on patterns of parasitism in 630 eastern chimpanzees (Pan troglodytes schweinfurthii) in Gombe National Park, Tanzania. 631 Am. J. Phys. Anthropol. 143, 534–544 (2010). 632 95. Howells, M. E., Pruetz, J. & Gillespie, T. R. Patterns of gastro-intestinal parasites and 633 commensals as an index of population and ecosystem health: the case of sympatric 634 Western chimpanzees (Pan troglodytes verus) and Guinea baboons (Papio hamadryas 635 papio) at Fongoli, Senegal. Am. J. Primatol. 73, 173–179 (2011). 636 96. Messenger, A. M., Barnes, A. N. & Gray, G. C. Reverse zoonotic disease transmission 637 (zooanthroponosis): a systematic review of seldom-documented human biological threats 638 to animals. PloS One 9, e89055 (2014). 639 97. Sá, R. M. et al. Gastrointestinal symbionts of chimpanzees in Cantanhez National Park, 640 Guinea-Bissau with respect to habitat fragmentation. Am. J. Primatol. 75, 1032–1041 641 (2013). 642 98. Bordes, F., Morand, S. & Ricardo, G. Bat fly species richness in Neotropical bats: 643 correlations with host ecology and host brain. Oecologia 158, 109–116 (2008). 644 99. Ezenwa, V. O., Price, S. A., Altizer, S., Vitone, N. D. & Cook, K. C. Host traits and 645 parasite species richness in even and odd-toed hoofed mammals, Artiodactyla and 646 Perissodactyla. Oikos 115, 526–536 (2006). 647 100. Morand, S. & Poulin, R. Density, body mass and parasite species richness of 648 terrestrial mammals. Evol. Ecol. 12, 717–727 (1998). 649 101. Nunn, C. L., Altizer, S., Jones, K. E. & Sechrest, W. Comparative tests of parasite 650 species richness in primates. Am. Nat. 162, 597–614 (2003). 651

652

653

654

655

656

21 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

657 Table 1. Bootstrap estimate of host richness in each fragment of Anamalai hills, India

Fragment name Bootstrap estimate SE Sample size of host species richness Aliyar_dam 21.9 0.8 210 Akkamalai 18.9 0.7 199 Anaikundi 20.1 0.8 150 Andiparai 22.6 1 256 Attakatty 11.7 0.6 15 Iyerpadi 25.5 1 398 Karian_shola 25.3 0.9 244 Korangumudi 24.9 0.8 356 Monica_estate 20.4 0.9 124 Monomboly 26.6 1 181 Nirar_dam 24.2 0.9 167 Pannimedu 17.5 1 55 Puthuthottam 24.9 0.8 426 Sethumadai 18.1 0.9 65 Shekkalmudi 16.0 0.8 42 Sirukundra 17.9 0.8 127 Uralikal 11.5 0.5 41 Varagaliyar 21.6 1 162 Varattuparai 23.4 0.9 397 658

659

22 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

660

661 Table 2. Comparison between two different models to explain bootstrap estimate of 662 parasite taxon richness in Anamalai hills, India

Models K logLik AIC delta weight Plantation + Settlement + Livestock + 9 -667.54 1353.07 0 0.887 Host richness +Host body size Fragment size + Isolation index + Host 8 -670.59 1357.19 4.112 0.113 richness +Host body size 663

664

665

666

667

668

669 Table 3. Comparison between two different models to explain bootstrap estimate of host 670 species richness in Anamalai hills, India

Models K logLik AIC delta weight Plantation + Settlement + Livestock 5 -46.73 103.47 0 0.971 Fragment size + Isolation index 4 -51.25 110.5 7.029 0.029 671

672

673

23 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

674 Table 4. Parasite taxa that were found only in disturbed or undisturbed fragments in 675 Anamalai hills, India. Highlighted parasite taxa were found in the corresponding fragment 676 group exclusively. 1. Current study; 2. Natural History Museum parasite database, London, 677 UK

Parasite taxa Parasite Family In livestock Known group samples1 human case2 Plantation only Baylisascaris sp. Nematodes Ascaridoidea Present present Nematodirus sp. Nematodes Trichostrongyloidea Present present Enterobius sp. Nematodes Oxyuroidea Present present Dictyocaulus sp. Nematodes Trichostrongyloidea Absent absent Uncinaria sp. Nematodes Ancylostomatoidea Absent present Schistosoma sp. Trematodes Schistosomatidae Absent present Metastrongylus sp. Nematodes Metastrongyloidea Absent present Clonorchis sp. Trematodes Opisthorchiidae Present present Toxoplasma sp. Apicomplexa Present present Isospora sp. Apicomplexa Present present Paragonimus sp. Trematodes Paragonimidae Present present Dipylidium sp. Cestodes Dilepididae Present present

Livestock presence only Dictyocaulus sp. Nematodes Trichostrongyloidea Absent absent Taenia sp. Cestodes Taeniidae Absent present Metastrongylus sp. Nematodes Metastrongyloidea Absent present Undisturbed (Livestock) only Paragonimus Trematodes Paragonimidae Present present

Settlement only Dictyocaulus sp. Nematodes Trichostrongyloidea Absent absent Uncinaria sp. Nematodes Ancylostomatoidea Absent present Metastrongylus sp. Nematodes Metastrongyloidea Absent present Undisturbed (Settlement) only Sarcocystis sp. Apicomplexa Present present 678

679

680

24 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

681

682 Figure 1. Map of Anamalai hills, Western Ghats, India with numbered study fragments. (1) 683 Aliyar dam, 2) Akkamalai, 3) Anaikundi, 4) Andiparai, 5) Attakatty, 6) Iyerpadi 7) 684 Karian_shola, 8) Korangumudi, 9) Monica_estate, 10) Monomboly, 11) Nirar_dam, 12) 685 Pannimedu, 13) Puthuthottam, 14) Sethumadai 15) Shekkalmudi 16) Sirukundra 17) Uralikal, 686 18) Varagaliyar and 19) Varattuparai WLS: Wildlife sanctuary; RF: Reserve Forest; NP: 687 National Park

688

689

25 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

690

691 Figure 2. Unstandardized effect size of predictor variables on bootstrap estimate of parasite 692 taxon richness in rainforest fragments of Anamalai hills, India. Estimates were plotted to 693 scale. Intercepts were omitted to avoid distortion of scale. Disturbance model was the best 694 fitted model based on AIC. Confidence intervals are represented by the lines around the 695 points— thick (α = 0.10) and thin (α = 0.05).

696

697

26 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

698

699 Figure 3. Unstandardized effect size of predictor variables on bootstrap estimate of host 700 species richness in rainforest fragments of Anamalai hills, India. Estimates were plotted to 701 scale. Intercepts were omitted to avoid distortion of scale. Disturbance model was the best 702 fitted model based on AIC. Confidence intervals are represented by the lines around the 703 points— thick (α = 0.10) and thin (α = 0.05). Host sample sizes are given in Table 1.

704

27 bioRxiv preprint doi: https://doi.org/10.1101/645044; this version posted May 22, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

705

706 Figure 4. Host community heterogeneity between undisturbed (absent) and disturbed 707 (present) rainforest fragments of Anamalai hills, India. Community heterogeneity is the 708 within group dispersion values based on Jaccard distance for presence/absence data.

709

28